Resumo

PARDO, Constanza; DE VRIES, Esther; ACERO, Diana  e  MURILLO, Raúl. Surveillance of overall cancer survival in Colombia using routine registries. rev.colomb.cancerol. [online]. 2015, vol.19, n.2, pp.81-99. ISSN 0123-9015.  https://doi.org/10.10167j.rccan.2015.03.001.

Objective: To evaluate surveillance methods of overall cancer survival based on secondary sources. Methods: An analysis was performed on data from patients with invasive colorectal cancer in two fixed cohorts of the National Cancer Institute (INC) (2007 and 2010). Government sources and two index dates (admission and first diagnosis) were used, as well as three methods to determine the date of death: estimated date by the mean of the difference the certified date and the date the death was registered (method 1): date estimated by the median of the difference (method 2); and exact date from the death certificate (method 3). The overall survival was estimated at two years for both cohorts, and at five years for the 2007 cohort, using the Kaplan-Meier method. Results: A total of 616 cases were analyzed. The survival estimated with the first diagnosis date was the highest. The survival at two years for the 2007 cohort was 57, 58, and 78% (methods 1, 2 and 3, respectively). For the 2010 cohort, the survival at two years was 61, 62, and 83% with the methods in their order. The survival at five years (2007 cohort) was 35% with methods 1 and 2, and 72% with method 3. Conclusions: The secondary sources available in Colombia provide useful information for estimating overall survival and allow generating a population base less bias in the date of last contact. It is necessary to improve the registry of clinical stage to improve comparability between hospital institutions. Using only the mortality information available in the hospital registries leads to an over-estimation of survival.

Palavras-chave : Neoplasm; Survival; Hospital-based registry; Mortality; Surveillance; Colombia.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )